Imaging based dosimetry for individualized internal emitter therapy Summary/Abstract The current standard practice for internal emitter therapy rarely involves pre-treatment planning to optimize the radiation absorbed dose to the tumor while avoiding critical organ (typically bone marrow) toxicity. This is in stark contrast to planning for external beam therapy where precise absorbed dose calculations to tumor and surrounding organs are now mandatory. Such therapy optimization has not been used in therapies such as radioimmunotherapy (RIT) primarily because accurate internal emitter dose estimation is relatively difficult and because clinical studies thus far have failed to show a consistent relationship between radiation absorbed dose and effect. These past studies typically relied on suboptimal methods of dose estimation and/or did not account for biologic factors that are also expected to affect outcome. The long-term goal is clinical implementation of effective individualized treatment planning for radionuclide therapies such as I-131 tositumomab RIT in non- Hodgkin's lymphoma (NHL). The objective in the present application is to develop accurate imaging based methods for tumor and bone marrow dosimetry and to develop predictive models for tumor response and bone marrow toxicity incorporating key dosimetric factors as well as biologic factors, such as differential proliferation, radiosensitivity and sensitivity to the unlabeled antibody, that will be determined by biomarker studies. The central hypothesis is that integrating dosimetry and biology will enable better prediction of RIT outcome than is obtained with the strictly dosimetric measures commonly used. The hypothesis will be tested in refractory NHL patients undergoing I-131 RIT in standard clinical care and in frontline patients on a phase II clinical trial. To accomplish the objective of this application the following specific aims will be pursued: 1) Develop SPECT reconstruction methods that utilize CT-image information (without needing explicit segmentation of target boundaries) to more accurately estimate the 3-D activity distribution in targets, because dose heterogeneity impacts the effect of the therapy; 2) Develop imaging based bone marrow dosimetry coupling SPECT/CT with Monte Carlo radiation transport and accounting for variations in marrow composition as determined by quantitative CT; 3) Using patient data develop a multivariate regression model for predicting therapy outcome (response, toxicity) based on dosimetric factors and biologic factors from biomarker studies (e.g., Ki-67, p53 and FLT3-L from immunohistochemistry/immunoassay); and 4) Develop a mechanistic model (with and without modification for low dose hyper-radiosensitivity) to determine the equivalent biologic effect for predicting therapy outcome. The proposed work is innovative because unlike past studies focusing purely on dosimetry this work combines dosimetric factors with biologic factors to arrive at the optimal model for individualized treatment planning. The contribution is highly significant because once the methodologies and predictive models for treatment optimization are established, they will be used in the future by physicians for patient selection and to tailor RIT on a patient-by-patient basis to considerably improve the efficacy of the treatment. PUBLIC HEALTH RELEVANCE: The proposed research is relevant to public health because it is expected to provide the clinicians with the methodology and predictive models essential for future radioimmunotherapy (RIT) treatment optimization based on accurate pre-therapy dosimetric calculations and results from widely available biomarkers. The efficacy of RIT is likely to improve substantially with such individualized planning, hence the proposed research is relevant to the part of NIH's mission that pertains to developing fundamental knowledge and the application of that knowledge to extend healthy life and reduce the burdens of illness and disability.